Background: Core temperature measurements are an important component of perioperative patient monitoring. It is fairly easy to obtain core temperature measurements invasively in anesthetized patients. However, such measurements are more difficult to obtain noninvasively in awake patients. Recently, a new version of a temporal artery thermometer for noninvasive core temperature measurements (TemporalScanner TAT-5000) was introduced with accuracy and precision advertised as being comparable to invasive core temperature measurements. In this study, we sought to determine if this new thermometer is an acceptable substitute for invasive bladder temperature measurement.
Methods: In 35 patients undergoing neurosurgical interventions and 35 patients in the neurosurgical intensive care unit, measurements from the temporal artery thermometer were compared with those from a bladder thermometer. Four measurements were obtained from each patient.
Results: Overall 280 measurement pairs were obtained. The mean bias between the methods was 0.07 degrees C +/- 0.79 degrees C; the limits of agreement were approximately 3 times greater than the a priori defined limit of +/-0.5 degrees C (-1.48 to 1.62). The sensitivity for detecting fever (core temperature >37.8 degrees C) using the temporal artery thermometer was 0.72, and the specificity was 0.97. The positive predictive value for fever was 0.89; the negative predictive value was 0.94. The sensitivity for detecting hypothermia (core temperature <35.5 degrees C) was 0.29, and the specificity was 0.95. The positive predictive value for hypothermia was 0.31, and the negative predictive value was 0.95.
Conclusions: The results of this study do not support the use of temporal artery thermometry for perioperative core temperature monitoring; the temporal artery thermometer does not provide information that is an adequate substitute for core temperature measurement by a bladder thermometer.